CN115481918A - Active sensing and predictive analysis system for unit state based on source network load storage - Google Patents
Active sensing and predictive analysis system for unit state based on source network load storage Download PDFInfo
- Publication number
- CN115481918A CN115481918A CN202211197518.XA CN202211197518A CN115481918A CN 115481918 A CN115481918 A CN 115481918A CN 202211197518 A CN202211197518 A CN 202211197518A CN 115481918 A CN115481918 A CN 115481918A
- Authority
- CN
- China
- Prior art keywords
- data
- power
- prediction
- load
- historical
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004458 analytical method Methods 0.000 title claims abstract description 115
- 238000003860 storage Methods 0.000 title claims abstract description 36
- 238000000034 method Methods 0.000 claims abstract description 42
- 230000010354 integration Effects 0.000 claims abstract description 30
- 238000005057 refrigeration Methods 0.000 claims description 78
- 238000010248 power generation Methods 0.000 claims description 44
- 238000012545 processing Methods 0.000 claims description 31
- 238000004364 calculation method Methods 0.000 claims description 28
- 238000012423 maintenance Methods 0.000 claims description 25
- RYAUSSKQMZRMAI-YESZJQIVSA-N (S)-fenpropimorph Chemical compound C([C@@H](C)CC=1C=CC(=CC=1)C(C)(C)C)N1C[C@H](C)O[C@H](C)C1 RYAUSSKQMZRMAI-YESZJQIVSA-N 0.000 claims description 24
- 238000004519 manufacturing process Methods 0.000 claims description 23
- 230000032683 aging Effects 0.000 claims description 20
- 238000005516 engineering process Methods 0.000 claims description 15
- 238000012544 monitoring process Methods 0.000 claims description 13
- 241000668842 Lepidosaphes gloverii Species 0.000 claims description 11
- 238000012351 Integrated analysis Methods 0.000 claims description 6
- 238000007599 discharging Methods 0.000 claims description 6
- 230000007613 environmental effect Effects 0.000 claims description 4
- 230000000694 effects Effects 0.000 abstract description 8
- 238000013135 deep learning Methods 0.000 description 16
- 238000007405 data analysis Methods 0.000 description 13
- 238000005457 optimization Methods 0.000 description 9
- 238000012360 testing method Methods 0.000 description 6
- 230000007774 longterm Effects 0.000 description 5
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 5
- 238000012937 correction Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000005684 electric field Effects 0.000 description 3
- 230000000737 periodic effect Effects 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000003442 weekly effect Effects 0.000 description 2
- 238000010521 absorption reaction Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 229910052799 carbon Inorganic materials 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06312—Adjustment or analysis of established resource schedule, e.g. resource or task levelling, or dynamic rescheduling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Marketing (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Educational Administration (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a unit state active sensing and predictive analysis system and method based on source network load storage, wherein the system comprises a new energy layer module, a computing layer module, a load layer module and an integration module; the new energy layer module is used for acquiring a first parameter group and determining first prediction data according to the first parameter group; the computing layer module is used for acquiring a second parameter group and determining second prediction data according to the second parameter group; the load layer module is used for acquiring a third parameter group and determining third prediction data according to the second parameter group, the second prediction data and the third parameter group; and the integration module is used for integrating the first prediction data, the second prediction data and the third prediction data and determining integration information, and the integration information is used for determining the optimized scheduling of the source network load storage unit. The scheduling effect of the source network load storage unit can be improved through the characteristics of the embodiment of the invention.
Description
Technical Field
The invention relates to the technical field of electric power, in particular to a unit state active sensing and predictive analysis system based on source network load storage.
Background
Under the double-carbon target, the integration of source network load and storage also becomes an important hotspot, wherein, the integration of source network load and storage is compared with the traditional power grid, the power grid development of the novel power system forms a pattern that the large power grid is dominant and various power grid forms are merged, so that a multi-level micro-power grid can be formed in different areas such as families, communities, parks and the like, and the problems of large-scale new energy and novel load access and plug-and-play are solved. A unidirectional process of ' generation-transmission-transformation-distribution-use ' of a traditional power system is formed into a source-network-load-storage ' integrated cyclic process, and the power generation and absorption ratio of new energy is improved.
In the prior art, the management of important units in the source network load and storage integration is generally directly adjusted through the existing data, so that the actual mapping relation between the computing power and the electric power is unclear, and the scheduling effect of the units in the source network load and storage is poor.
Disclosure of Invention
The embodiment of the invention provides a unit state active sensing and predictive analysis system and method based on source network load storage, and aims to solve the problem of poor scheduling effect of a unit in source network load storage in the prior art.
The embodiment of the invention provides a unit state active sensing and predictive analysis system based on source network load storage, which comprises a new energy layer module, a computing layer module, a load layer module and an integration module, wherein the new energy layer module comprises a new energy layer module, a computing layer module and a load layer module;
the new energy layer module is configured to obtain a first parameter set, and determine first prediction data according to the first parameter set, where the first prediction data includes at least one of the following: the power generation power of the wind-light combined station and the power generation power of the wind-light combined cluster;
wherein the first parameter set comprises at least one of: meteorological data, power generation attenuation rate and operation and maintenance parameters;
the computation layer module is configured to obtain a second parameter set, and determine second prediction data according to the second parameter set, where the second prediction data includes at least one of the following data: the total core number of a virtual server needed by the computing task corresponding to a Central Processing Unit (CPU) and the load rate of a physical server;
wherein the second parameter set comprises at least one of: historical data of the virtual server required by the computing task corresponding to the total core number of the Central Processing Unit (CPU) and historical data of the load rate of the physical server;
the load layer module is configured to obtain a third parameter group, and determine third prediction data according to the second parameter group, the second prediction data, and the third parameter group, where the third prediction data includes at least one of the following: calculating the corresponding prediction data of the electric load required by the power task, the refrigeration facility and the charging facility;
wherein the third parameter set comprises at least one of: historical data of power loads, running state parameters, equipment aging rates, ambient temperatures, charging facility power supply loads and charging and discharging attenuation rates input by a server refrigeration facility, a machine room refrigeration facility and other production and living refrigeration facilities;
the integration module is used for integrating the first prediction data, the second prediction data and the third prediction data and determining integration information, and the integration information is used for determining optimized scheduling of the source network load and storage unit.
The embodiment of the invention also provides a unit state active sensing and predictive analysis method based on source network load storage, which comprises the following steps:
acquiring a first parameter group, and determining first prediction data according to the first parameter group, wherein the first prediction data comprises at least one of the following items: the power generation power of the wind-light combined station and the power generation power of the wind-light combined cluster;
wherein the first parameter set comprises at least one of: meteorological data, power generation attenuation rate and operation and maintenance parameters;
acquiring a second parameter group, and determining second prediction data according to the second parameter group, wherein the second prediction data comprises at least one of the following items: the total core number of a virtual server needed by the computing task corresponding to a Central Processing Unit (CPU) and the load rate of a physical server;
wherein the second parameter set comprises at least one of: historical data of the virtual server corresponding to the CPU total core number of the central processing unit and historical data of the load rate of the physical server required by the computing task;
acquiring a third parameter group, and determining third prediction data according to the second parameter group, the second prediction data and the third parameter group, wherein the third prediction data comprises at least one of the following data: calculating the corresponding prediction data of the electric load required by the power task, the refrigeration facility and the charging facility;
wherein the third parameter set includes at least one of: historical data of power loads, running state parameters, equipment aging rates, ambient temperatures, charging facility power supply loads and charging and discharging attenuation rates input by a server refrigeration facility, a machine room refrigeration facility and other production and living refrigeration facilities;
and integrating the first prediction data, the second prediction data and the third prediction data and determining integration information, wherein the integration information is used for determining optimized scheduling of the source network load and storage unit.
In the embodiment of the invention, the unit state active sensing and prediction analysis system firstly obtains a first parameter group through a new energy layer module, determines first prediction data according to the first parameter group, similarly obtains a second parameter group and a third parameter group through a calculation layer module and a load layer module, determines second prediction data according to the second parameter group, and determines third prediction data according to the third parameter group, wherein the first prediction data, the second prediction data and the third prediction data respectively correspond to a new energy layer, a calculation layer and a load layer, and finally determines integration information on the basis of the prediction data through an integration module, so that basic support information is provided for optimizing cooperative scheduling calculation and electric power, and the scheduling effect on a source network load storage unit is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without inventive labor.
FIG. 1 is a schematic structural diagram of an active sensing and predictive analysis system according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart of an active sensing and predictive analysis system according to an embodiment of the present invention;
FIG. 3 is a second schematic flow chart of an active sensing and predictive analysis system according to an embodiment of the present invention;
FIG. 4 is a third schematic flow chart of an active sensing and predictive analysis system according to an embodiment of the present invention;
FIG. 5 is a fourth schematic flowchart of an active sensing and predictive analysis system according to an embodiment of the present invention;
FIG. 6 is a fifth flowchart of an active sensing and predictive analysis system according to an embodiment of the present invention;
FIG. 7 is a sixth flowchart of an active sensing and predictive analysis system according to an embodiment of the present invention;
FIG. 8 is a seventh schematic flow chart of an active sensing and predictive analysis system according to an embodiment of the present invention;
FIG. 9 is a flowchart illustrating a method for actively sensing and predicting a cell state according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The terms "first," "second," and the like in the embodiments of the present invention are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
Referring to fig. 1, fig. 1 is a schematic structural diagram of a system for actively sensing and predicting and analyzing a unit state based on source network load and storage according to an embodiment of the present invention, and as shown in fig. 1, the system may include a new energy layer module 10, an effort layer module 20, a load layer module 30, and an integration module 40;
the new energy layer module 10 is configured to obtain a first parameter set, and determine first prediction data according to the first parameter set, where the first prediction data includes at least one of the following: the power generation power of the wind-light combined station and the power generation power of the wind-light combined cluster;
wherein the first parameter set comprises at least one of: meteorological data, power generation attenuation rate and operation and maintenance parameters;
the computation layer module 20 is configured to obtain a second parameter set, and determine second prediction data according to the second parameter set, where the second prediction data includes at least one of the following: the total core number of a virtual server needed by the computing task corresponding to a Central Processing Unit (CPU) and the load rate of a physical server;
wherein the second parameter set comprises at least one of: historical data of the virtual server corresponding to the CPU total core number of the central processing unit and historical data of the load rate of the physical server required by the computing task;
the load layer module 30 is configured to obtain a third parameter group, and determine third prediction data according to the second parameter group, the second prediction data, and the third parameter group, where the third prediction data includes at least one of the following data: calculating the corresponding prediction data of the electric load required by the power task, the refrigeration facility and the charging facility;
wherein the third parameter set includes at least one of: historical data of power loads, running state parameters, equipment aging rate, ambient temperature, power supply loads of charging facilities and charge-discharge attenuation rates input by server refrigeration facilities, machine room refrigeration facilities and other production and living refrigeration facilities;
an integration module 40, configured to integrate the first prediction data, the second prediction data, and the third prediction data and determine integration information, where the integration information is used to determine an optimized scheduling for a source network load-store unit.
In this embodiment, in the embodiment of the present invention, the unit state active sensing and prediction analysis system first obtains a first parameter group through the new energy layer module 10, and determines first prediction data according to the first parameter group, and similarly, obtains a second parameter group and a third parameter group through the computation layer module 20 and the load layer module 30, and determines second prediction data according to the second parameter group, and determines third prediction data according to the third parameter group, where the first prediction data, the second prediction data, and the third prediction data respectively correspond to the new energy layer, the computation layer, and the load layer, and finally, the unit state active sensing and prediction analysis system determines integration information on the basis of the prediction data through the integration module 40, so as to provide basic support information for optimizing cooperative scheduling computation and power, thereby improving the scheduling effect on the unit in the source network load storage.
It can be understood that the new energy layer corresponding to the new energy layer module 10 may represent a green power source, and the new energy layer may provide power prediction data of ultra-short term, and medium-long term output of the wind power and photovoltaic plant station for the source grid load storage integrated project, and provide basic information support for developing source grid load storage integrated optimization scheduling operation modes across seasons, months, days, and real-time, and optimization decisions of power market trading.
Specifically, the first prediction data determined by the new energy layer mainly includes wind power and photovoltaic power generation power on the scale of year, month, week, day, hour and minute, and in addition, the data source of the first parameter group may include two parts: firstly, weather history and forecast data input by an external weather database; and the other is historical output power data and electric field planning operation and maintenance measures input by the wind and light combined station. And obtaining a multi-time scale output power predicted value, namely the first predicted data, according to the first parameter group.
In addition, the computing power layer corresponding to the computing power layer module 20 may represent a computing power load state of a data center server, provide medium and long term computing power task state prediction data for a source network load storage integrated project based on a big data center, and provide basic data support for developing important load prediction analysis and computing power and power collaborative optimization scheduling.
Specifically, the second prediction data determined by the computing layer module 20 may mainly include monthly and yearly prediction data of the total number of CPU cores of the CPU of the virtual servers required by a single specific computing task, all computing tasks, and the load rate of the physical server, and the source of the second parameter group may include: the method comprises the steps of obtaining historical data such as the total core number of a Central Processing Unit (CPU) of a virtual server, the load rate of a physical server and the like required by a single periodic and long-term computing task and all computing tasks, and computing requirements to be newly added by a user in the future. The second parameter set may be monthly and yearly forecast data for obtaining the total number of CPU cores of the CPU of the virtual server and the load rate of the physical server required for a single specific computing task and all computing tasks, that is, the second forecast data.
In addition, the load layer corresponding to the load layer module 30 may represent an important load type of the big data center, and provide ultra-short-term, and medium-long-term operation power prediction data of the important load for the source network load storage integration project based on the big data center, and provide basic data and technical support for developing flexible optimization scheduling of the important load such as data center internet technology equipment, refrigeration facilities, charging facilities, and the like.
Specifically, the third prediction data determined by the load layer module 30 mainly includes prediction data of a single power calculation task, and year, month, week, day, hour scale and minute scale of power consumption required by all power calculation tasks, and in addition, a basic data source corresponding to the third prediction data may include the following two parts:
(1) And historical and test data, including historical data of the total number of the Central Processing Units (CPU) of the virtual servers required by each power calculation task and all the power calculation tasks, historical data of the total power load of the servers, and the total number of the CPU of the virtual servers required by a single power calculation task and the test data of the power loads of the corresponding servers.
(2) And reporting and predicting data in real time, wherein the data comprises short-term computing power task allocation results and the total number of Central Processing Units (CPUs) of the virtual servers required by a single specific task and all computing power tasks.
Optionally, the new energy layer module comprises a long-scale high-precision power prediction analysis unit of the wind-solar combined station and a centralized power monitoring analysis and high-precision prediction unit of a regional multi-wind-field station cluster;
the wind and light combined station long-scale high-precision power prediction analysis unit is used for fitting the first parameter set to obtain first fitting information, and predicting the power generation power of a single wind and light combined station according to the first fitting information;
and the centralized power monitoring analysis and high-precision prediction unit of the regional multi-wind-field station cluster is used for fitting the first parameter group to obtain second fitting information, and predicting the power generation power of the wind-light combined cluster according to the second fitting information.
In this embodiment, the new energy layer module may include a wind and light combined plant long-length high-precision power prediction analysis unit and a centralized power monitoring analysis and high-precision prediction unit of a regional multi-wind-field-station cluster, where the wind and light combined plant long-length high-precision power prediction analysis unit and the centralized power monitoring analysis and high-precision prediction unit of the regional multi-wind-field-station cluster are respectively used to predict and obtain the generated power of the wind and light combined cluster of the generated power of a single wind and light combined plant, so as to improve the scheduling effect on the source network load storage unit.
Specifically, the long-scale high-precision power prediction unit of the wind and light combined station is mainly used for predicting the annual, monthly, weekly, daily, hourly and minute-scale power generation power of a single wind and light combined station, and prediction data are used for source network load storage integrated comprehensive optimization scheduling and power market trading optimization decision; similarly, the centralized power monitoring analysis and high-precision prediction unit of the regional multi-wind-light-field station cluster is mainly used for predicting annual, monthly, weekly, daily, hourly and minute-scale generation power of a plurality of wind-light combined clusters, and prediction data are used for source network, load and storage integrated comprehensive optimization scheduling and power market trading optimization decision.
Optionally, the first fitting information corresponding to the wind-solar combined station long-scale high-precision power prediction analysis subsystem is as follows:
wherein, y a Historical and predictive data, x, representing the total generated electrical power of a wind farm a Historical and predicted data, x 'representing different meteorological parameters' a Historical and predicted data, x ", representing different operation and maintenance parameters a Historical and predictive data representing the rate of decay of power generation by the plant.
Referring to fig. 2, fig. 2 is a schematic flow chart corresponding to the long-scale high-precision power prediction analysis subsystem of the wind-light combined plant, and as shown in fig. 2, the data source of the long-scale high-precision power prediction analysis subsystem of the wind-light combined plant includes two parts: firstly, the maximum 15-minute resolution is required to be achieved by meteorological history and forecast data input by an external meteorological database, and the maximum 15-minute resolution comprises multi-altitude layer wind speed/direction, air pressure, irradiation, temperature, solar incident angle, solar azimuth angle and the like; the second is historical output power data and electric field planning operation and maintenance measures input by the wind and light combined station, for example: maintenance time, number of maintenance equipment, etc.
Under the condition of considering the power generation attenuation rate and operation and maintenance measures of equipment, based on historical data, fitting equations between wind and light combined power generation power and meteorological data, power generation attenuation rate and operation and maintenance parameters are constructed by means of digital twin and big data analysis and the like, then prediction equations between the wind and light combined power generation power and the meteorological data, the power generation attenuation rate and the operation and maintenance parameters are constructed by using methods such as deep learning, digital twin and the like according to meteorological parameters, planned operation and maintenance measures and power generation attenuation rate prediction data of different time scales on the basis of the fitting equations, the two equations are automatically updated and adjusted according to the deviation between the prediction data and actual data, and the prediction accuracy is increased along with the expansion of a data set.
The annual and monthly predicted output power is mainly based on historical data of the current month of many years, the updating frequency of the prediction equation is small, and the updating frequency of the prediction equation in the daily scale, the hour scale and the minute scale is increased.
The first fitting information can be expressed as a fitting equation, and the fitting equation is processed by methods such as deep learning and digital twinning to obtain a prediction equation as follows:
wherein, y a Historical and predictive data, x, representing the total generated electrical power of a wind farm a Historical and predicted data, x 'representing different meteorological parameters' a Historical and predicted data, x ″, representing different operation and maintenance parameters a Historical and predictive data representing the rate of decay of power generation by the plant,
representing the deviation of the predicted data from the actual data.
It should be noted that the first fitting information, for example, the fitting equation, corresponding to the wind and light combined plant long-scale high-precision power prediction analysis subsystem may be an equation artificially set by a digital twin or large data analysis method according to input data, and then the set equation is used to obtain the prediction equation corresponding to the wind and light combined plant long-scale high-precision power prediction analysis subsystem by a digital twin or deep learning method, which is not limited in the embodiment of the present invention.
Optionally, the second fitting information corresponding to the centralized power monitoring analysis and high-precision prediction subsystem of the regional multi-wind field station cluster is as follows:
wherein, y b Historical and predictive data, x, representing the total generated electrical power of a wind-solar cluster b Representing history and predictions of different meteorological parametersMeasurement data, x' b Historical and predicted data, x ″, representing different operation and maintenance parameters b Historical period and expected life data, x ″, representative of different corporate sites' b Historical and predictive data representing the rate of decay of power generation by the plant.
Referring to fig. 3, fig. 3 is a schematic flow chart corresponding to the centralized power monitoring analysis and high-precision prediction subsystem of the regional multi-wind farm station cluster, and as shown in fig. 3, data of the centralized power monitoring analysis and high-precision prediction subsystem of the regional multi-wind farm station cluster may be from two parts, for example: the method comprises the steps that firstly, meteorological history and forecast data input by an external meteorological database comprise multi-altitude layer wind speed/direction, air pressure, irradiation, temperature, solar incident angle, solar azimuth angle and the like; the historical output power data input by the wind-solar power generation cluster and the electric field planned operation and maintenance measures, such as: such as overhaul time, overhaul equipment number, etc.
Under the condition of considering equipment power generation attenuation rate, equipment life difference of different united stations and operation and maintenance measures, based on historical data, fitting equations between total power generation power and meteorological data, power generation attenuation rate and operation and maintenance parameters of regional multi-wind-field station clusters are constructed by methods such as digital twinning and big data analysis, then prediction equations between the total power generation power and the meteorological data, the power generation attenuation rate and the operation and maintenance parameters of the multi-wind-field station clusters are constructed by methods such as deep learning and digital twinning on the basis of the fitting equations according to the meteorological parameters, planned operation and maintenance measures, power generation attenuation rate prediction data and the expected life of different united stations, and the prediction equations are automatically updated and adjusted according to the deviation of the prediction data and actual data.
The annual and monthly predicted output power is mainly based on historical data of the current month of many years, the updating frequency of the prediction equation is small, and the updating frequency of the prediction equation in the daily scale, the hour scale and the minute scale is increased.
The second fitting information may be a fitting equation, and the fitting equation is subjected to methods such as deep learning and digital twinning to obtain a prediction equation as follows:
wherein, y b Historical and predictive data, x, representing the total generated electrical power of a wind-solar cluster b Historical and predicted data, x 'representing different meteorological parameters' b Historical and predicted data, x ", representing different operation and maintenance parameters b Historical period and expected life data, x 'representative of different corporate stations' b Historical and predictive data representing the rate of decay of power generation by the plant,
representing the deviation between the last predicted data and the actual result.
It should be noted that the second fitting information, for example, the fitting equation, corresponding to the centralized power monitoring analysis and high-precision prediction subsystem of the regional windward field station cluster may be an equation artificially set by a digital twinning or big data analysis method according to input data, and then the set equation is used to obtain the prediction equation of the centralized power monitoring analysis and high-precision prediction subsystem corresponding to the regional windward field station cluster by a digital twinning or deep learning method, which is not limited in the embodiment of the present invention.
Optionally, the computing power layer module comprises a data center computing power task prediction analysis unit;
and the data center computing power task prediction analysis unit is used for fitting the second parameter group to obtain third fitting information, and determining the second prediction data according to the third fitting information.
Referring to fig. 4, fig. 4 is a schematic flow chart corresponding to the data center computing power task prediction analysis unit, and as shown in fig. 4, the data source of the data center computing power task prediction analysis unit may be historical data of the total number of cores of the central processing units CPU of the single periodic and long-term computing power task and the virtual server required by all computing power tasks and historical data of the load rate of the physical server, which are input by the computing power scheduling center.
Then, a fitting equation of the total core number of the Central Processing Units (CPUs) of the virtual servers and the load rate of the physical servers, which is required by a single specific task and all the computing power tasks, may be constructed by means of big data analysis and the like, that is, the third fitting information, which may be represented by the following equation:
y c ＝f(t c )
y d ＝f(t d )
y e ＝f(t e )
wherein y is c 、y d 、y e Historical and predicted data representing the total number of CPU cores of the central processing unit of the virtual server required by a single specific task and all computing tasks and the load rate of the physical server respectively, t c 、t d 、t e Respectively representing the historical and predicted operation time periods, x, corresponding to a single specific task, all computing tasks and the whole physical server c 、x d 、x e The total number of the central processing unit CPU cores of the virtual server to be newly added in the future by the user corresponding to a single specific task and all the calculation tasks and the physical server to be newly added in the future of the whole data center are represented respectively.
Under the condition that computing power requirements to be newly added of related users are considered, methods such as deep learning are used for constructing a Central Processing Unit (CPU) total core number of a virtual server and a next month and next year prediction equation of a physical server load rate, wherein the CPU total core number and the next month and the next year prediction equation are needed by a single specific task and all computing power tasks, and the prediction equation is automatically updated and adjusted according to the deviation of prediction data and actual data.
Wherein, the above-mentioned prediction equation can be expressed as:
wherein y is c 、y d 、y e Historical and predicted data representing the total number of CPU cores of the central processing units of the virtual servers required by a single specific task and all computing power tasks and the load rate of the physical servers respectively, t c 、t d 、t e Respectively representing the historical and predicted operation time periods, x, corresponding to a single specific task, all computing tasks and the whole physical server c 、x d 、x e Respectively representing the CPU total core number of the central processing unit of the virtual server to be newly added in the future by a user corresponding to a single specific task, all computing power tasks and the physical server to be newly added in the future of the whole data center,
representing the deviation between the last predicted data and the actual result.
It should be noted that, the third fitting information corresponding to the data center computing power task prediction analysis unit, for example, the fitting equation, may be an equation artificially set by a big data analysis method according to the input data, and then the set equation is used to obtain the prediction equation corresponding to the data center computing power task prediction analysis unit by a deep learning method, which is not limited in the embodiment of the present invention.
Optionally, the load layer module includes an internet technology equipment power load comprehensive analysis and prediction analysis unit, a data center refrigeration facility power load comprehensive analysis and prediction analysis unit, and a data center charging facility power load prediction analysis unit;
the internet technology equipment power load comprehensive analysis and prediction analysis unit is used for fitting the second prediction data and the second parameter group to obtain fourth fitting information, and determining the third prediction data according to the fourth fitting information;
the data center refrigeration facility power load comprehensive analysis and prediction analysis unit is used for fitting the second parameter group, the third parameter group and the second prediction data to obtain fifth fitting information, and determining prediction data of a machine room refrigeration facility and other production and living refrigeration facilities according to the fifth fitting information;
and the power load prediction analysis unit of the charging facility of the data center is used for fitting the third parameter group to obtain sixth fitting information, and determining the prediction data of the charging facility in the park according to the sixth fitting information.
In this embodiment, the load layer module may include an internet technology equipment power load comprehensive analysis and prediction analysis unit, a data center refrigeration facility power load comprehensive analysis and prediction analysis unit, and a data center charging facility power load prediction analysis unit, wherein the internet technology equipment power load comprehensive analysis and prediction analysis unit is configured to obtain prediction data of years, months, weeks, days, hours, and minutes of power loads required by a single power calculation task, all power calculation tasks, the data center refrigeration facility power load comprehensive analysis and prediction analysis unit is configured to obtain prediction data of years, months, weeks, days, hours, and minutes of a machine room refrigeration facility, other production and living refrigeration facilities, and the data center charging facility power load prediction analysis unit is configured to obtain prediction data of years, months, weeks, days, hours, and minutes of a charging facility in a campus, so as to provide basic data and technical support for flexible optimization scheduling of important loads of the data center internet technology equipment, refrigeration facility, and charging facility, and the like, thereby improving a scheduling effect on the source grid power storage unit.
Optionally, the fourth fitting information corresponding to the power load comprehensive analysis and prediction analysis subsystem of the internet technology equipment is as follows:
y f ＝f(x f )
y g ＝f(x g )
wherein y is f 、y g Historical and predicted data representing the power loads required by a single power calculation task and all power calculation tasks in a short term and a medium-long term respectively, x f 、x g Respectively representing short-term and medium-and long-term single computational tasksAnd historical and predicted data of the total number of cores of the central processing unit CPU of the virtual server required by all the computing tasks.
Referring to fig. 5, fig. 5 is a schematic flow chart corresponding to the integrated analysis and predictive analysis subsystem for power load of the internet technology equipment, and as shown in fig. 5, a data source of the integrated analysis and predictive analysis subsystem for power load of the internet technology equipment may include two parts, for example:
(1) Historical and test data, namely historical data of the total core number of the Central Processing Units (CPUs) of the virtual servers required by each power calculation task and all the power calculation tasks of the servers and historical data of the total power load of the physical servers, which are input by the power calculation scheduling center, are subjected to field scientific tests to obtain the total core number of the CPUs of the virtual servers required by a single power calculation task and the test data of the power load of the corresponding servers.
(2) And reporting and predicting data in real time, wherein the data comprise short-term computing power task allocation results input by a computing power scheduling center, namely the total number of Central Processing Unit (CPU) cores of the virtual server required by the single computing power task and all computing power tasks, and the total number of CPU cores of the virtual server required by the single specific task and all computing power tasks input by the data central computing power task prediction analysis subsystem.
It should be noted that the power load comprehensive analysis and prediction analysis subsystem of the internet technology equipment can be divided into a fitting analysis and prediction part, for the fitting analysis part, based on history and test data, a fitting equation of the total number of Central Processing Units (CPUs) of the virtual servers required by all the power calculation tasks and a single power calculation task and the power load is respectively established by methods such as big data analysis and the like, and a history data set is updated every 15 minutes to drive correction and update of the fitting equation.
For the prediction part, two approaches are divided into short-term and medium-term:
(1) Short-term: based on real-time or early reporting of distribution results and corresponding fitting equations of single and all computing power tasks in a short period, short-term prediction equations of the total core number of Central Processing Units (CPUs) of virtual servers required by the single computing power task and all computing power tasks and the power load are respectively constructed by methods such as deep learning, predicted values of the power load required by the single computing power task and all computing power tasks in the next 15 minutes, the next 1 hour and the next day are obtained, and the prediction equations are automatically updated and adjusted according to the deviation of prediction data and actual data.
(2) Medium-long term: based on the central processing unit CPU total core number next month and next year prediction data of the virtual server required by a single specific task and all the calculation tasks and corresponding fitting equations, respectively constructing the central processing unit CPU total core number of the virtual server required by the single calculation task and all the calculation tasks and the medium-long term prediction equations of the power load by methods of deep learning and the like to obtain the predicted values of the power load required by the single calculation task, the next month and next year of all the calculation tasks, and automatically updating and adjusting the prediction equations according to the deviation of the prediction data and the actual data.
The fourth fitting information may be expressed as a fitting equation, and the fitting equation is processed to obtain a prediction equation as follows:
wherein y is f 、y g Historical and predicted data x representing the power load required by a single power calculation task and all power calculation tasks under short term and medium and long term respectively f 、x g Respectively representing the historical and predicted data of the total core number of the central processing unit CPU of the virtual server required by the single computing power task in the short term and the medium-long term and all the computing power tasks,
respectively, representing the deviation between the last predicted data and the actual result.
It should be noted that, the fourth fitting information, for example, the fitting equation, corresponding to the power load comprehensive analysis and predictive analysis subsystem of the internet technology equipment may be an equation artificially set by a big data analysis method according to input data, and then the set equation is used to obtain the predictive equation corresponding to the power load comprehensive analysis and predictive analysis subsystem of the internet technology equipment by a deep learning method, which is not limited in the embodiment of the present invention.
Optionally, the fifth fitting information corresponding to the data center refrigeration facility power load comprehensive analysis and predictive analysis subsystem is as follows:
wherein, y h 、y i Respectively representing the operating state parameters of the refrigerating facilities in the machine room, the history and the prediction data of the power load, y j 、y k Respectively representing the operating state parameters of other refrigerating facilities for production and living, the history and forecast data of electric load, t h 、t j Respectively representing the history and prediction time periods, x, corresponding to the operating state parameters i 、x k Represents the historical and predicted data x 'of the equipment aging rate in the machine room refrigeration facility and other production and living refrigeration facilities respectively' i 、x′ k Respectively representing the history and prediction data, x ″, of the environmental temperature corresponding to the refrigeration facilities of the machine room and other refrigeration facilities for production and living i Historical and forecast data representing physical server load rates. Referring to fig. 6 and 7, the data source of the data center refrigeration facility power load analysis and prediction subsystem may includeTwo parts are as follows:
(1) Historical data including electrical loads input by machine room refrigeration facilities, other production and living refrigeration facilities, operating state parameters such as: the temperature of supply and return water, the aging rate of equipment, the load rate of a physical server input by the computing power dispatching center and the environmental temperature input by an external meteorological database.
(2) The multi-time scale prediction data comprises equipment aging rates input by a machine room refrigeration facility and other production and living refrigeration facilities, physical server load rates input by a data center computing power task prediction analysis subsystem and ambient temperatures input by an external meteorological database.
It should be noted that the data center refrigeration facility power load comprehensive analysis and prediction analysis subsystem can be divided into a fitting analysis and prediction part, and the fitting analysis part can be divided into the following two cases:
(1) For the machine room refrigeration facility, based on the refrigeration facility electrical load, the operating state parameters, such as: historical data of water supply and return temperature, equipment aging rate, physical server load rate and environment temperature are established, a fitting equation of the electric load and running state parameters (such as water supply and return temperature), the physical server load rate, the equipment aging rate, the environment temperature and a fitting equation of the running state parameters and time of the machine room refrigeration facility are established by methods such as big data analysis, and a historical data set is updated every 15 minutes to drive correction and update of the fitting equation.
(2) For other production and living refrigeration facilities, based on historical data of the power load, the running state parameters (such as water supply and return temperature) and the ambient temperature of the refrigeration facility, a fitting equation of the power load and the running state parameters (such as water supply and return temperature), the equipment aging rate and the ambient temperature of the other production and living refrigeration facilities and a fitting equation of the running state parameters and time are established by methods such as big data analysis, and a historical data set is updated every 15 minutes to drive the correction and the update of the fitting equation.
For the prediction part, the following two cases can be classified:
(1) For a machine room refrigeration facility, a physical server is used for predicting load rate, equipment aging rate, environment temperature and other prediction data and corresponding fitting equations, the prediction equations of the operation parameters and time of the refrigeration facility are respectively constructed by deep learning and other methods based on year, month, week, day, hour scale and minute scale, the prediction equations of the power load and operation state parameters of the refrigeration facility, the load rate of the physical server, the equipment aging rate and the environment temperature are firstly obtained, then the prediction data of the power load of the refrigeration facility are further obtained based on the known prediction data of the operation parameters, the equipment aging rate, the prediction load rate of the physical server, the environment temperature and the like, and the prediction equations are automatically updated and adjusted according to the deviation of the prediction data and the actual data.
(2) For other production and living refrigeration facilities, prediction data such as equipment aging rate and ambient temperature and corresponding fitting equations are used, prediction equations of refrigeration facility operation parameters and time are respectively constructed by deep learning and other methods based on year, month, week, day, hour scale and minute scale, operation parameter prediction data are firstly obtained for refrigeration facility power load and operation state parameters, equipment aging rate and ambient temperature, then the refrigeration facility power load prediction data are further obtained based on the known prediction data such as operation parameters, equipment aging rate and ambient temperature, and the prediction equations are automatically updated and adjusted according to the deviation of the prediction data and actual data.
The fifth fitting information may be expressed as a fitting equation, and the fitting equation is processed by the above procedure to obtain a prediction equation of the following formula:
wherein, y h 、y i Respectively representing operating state parameters of the machine room refrigeration facility, historical and predicted data of the electrical load, y j 、y k Respectively representing the operating state parameters of other production and living refrigeration facilities, the history and prediction data of the power load, t h 、t j Respectively representing the historical and predicted time periods, x, corresponding to the operating condition parameters i 、x k Represents the historical and predicted data x 'of the equipment aging rate in the machine room refrigeration facility and other production and living refrigeration facilities respectively' i 、x′ k Respectively representing history and prediction data x' of the environmental temperature corresponding to the machine room refrigeration facility and other production and living refrigeration facilities i Historical and forecast data representing physical server load rates,
respectively, representing the deviation between the last predicted data and the actual result.
It should be noted that, the fifth fitting information, for example, the fitting equation, corresponding to the data center refrigeration facility power load integrated analysis and predictive analysis subsystem may be an equation artificially set by a big data analysis method according to input data, and then the set equation is subjected to a deep learning method to obtain the predictive equation corresponding to the data center refrigeration facility power load integrated analysis and predictive analysis subsystem, which is not limited in the embodiment of the present invention.
Optionally, the sixth fitting information corresponding to the data center charging facility power load prediction analysis subsystem is as follows:
y l ＝f(t l )+f(x l )
wherein, y l Historical and forecast data representing charging facility power supply load, considering average highest power supply load of user, average power supply load of specific date, t l Representative charging deviceTime period of operation, x l History and prediction data representing charge and discharge decay rates.
Referring to fig. 8, fig. 8 is a flow chart corresponding to the data center charging facility power load prediction analysis subsystem, and as shown in fig. 8, the data source of the data center charging facility power load prediction analysis subsystem may be historical data including charging facility power supply load and charging and discharging attenuation rate, and prediction data including charging and discharging attenuation rate.
And then, a fitting equation of the power supply load of the charging facility, the charge-discharge attenuation rate and the time is constructed by means of big data analysis and the like, and overall user characteristics including a main charging time period of the user, the average highest power supply load of the user, the average power supply load of a specific date (weekend, holiday) and the like are deeply analyzed.
And finally, with the overall user characteristics as boundary conditions, using methods such as deep learning to construct a prediction equation of the power supply load, the charge and discharge attenuation rate and the time of the charging facility, and automatically updating and adjusting the prediction equation according to the deviation of the prediction data and the actual data.
The sixth fitting information may be a fitting equation, and the fitting equation is processed to obtain a prediction equation of the following formula:
wherein, y l Historical and forecast data representing charging facility power supply load, considering average highest power supply load of user, average power supply load of specific date, t l Representing charging facility operating time period, x l Historical and predicted data representing charge and discharge decay rates,
representing the deviation between the last predicted data and the actual result.
It should be noted that, the sixth fitting information, for example, the fitting equation, corresponding to the data center charging facility power load prediction analysis subsystem may be an equation artificially set by a big data analysis method according to the input data, and then the set equation is subjected to a deep learning method to obtain the prediction equation corresponding to the data center charging facility power load prediction analysis subsystem, which is not limited in this embodiment of the present invention.
The embodiment of the invention predicts the future running states of new energy, computing power and important loads under the condition of considering various parameters, thereby solving the practical problem that the state sensing and prediction of the source network load and storage integrated important unit based on a big data center is lacked at present, establishing the fitting and predicting relationship between the computing power and the electric power, further providing support information for optimizing and scheduling the electric power and the computing power for the source network load and storage integrated system, and improving the scheduling effect on the source network load and storage integrated unit.
The embodiment of the present invention further provides a method for actively sensing and predicting a unit state based on source network load and store, which is applied to a system for actively sensing, predicting and analyzing a unit state, please refer to fig. 9, where fig. 9 is a schematic flow diagram of the method for actively sensing and predicting a unit state based on source network load and store, and the method includes:
wherein the first parameter set comprises at least one of: meteorological data, power generation attenuation rate and operation and maintenance parameters;
wherein the second parameter set comprises at least one of: historical data of the virtual server corresponding to the CPU total core number of the central processing unit and historical data of the load rate of the physical server required by the computing task;
wherein the third parameter set comprises at least one of: historical data of power loads, running state parameters, equipment aging rate, ambient temperature, power supply loads of charging facilities and charge-discharge attenuation rates input by server refrigeration facilities, machine room refrigeration facilities and other production and living refrigeration facilities;
and 904, integrating the first prediction data, the second prediction data and the third prediction data and determining integration information, wherein the integration information is used for determining optimal scheduling of the source network load storage unit.
The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (10)
1. A unit state active sensing and predictive analysis system based on source network load storage is characterized by comprising a new energy layer module, a computing layer module, a load layer module and an integration module;
the new energy layer module is configured to obtain a first parameter group, and determine first prediction data according to the first parameter group, where the first prediction data includes at least one of: the power generation power of the wind-light combined station and the power generation power of the wind-light combined cluster;
wherein the first parameter set comprises at least one of: meteorological data, power generation attenuation rate and operation and maintenance parameters;
the computation layer module is configured to obtain a second parameter set, and determine second prediction data according to the second parameter set, where the second prediction data includes at least one of the following data: the total core number of a virtual server corresponding to a Central Processing Unit (CPU) and the load rate of a physical server required by the computing task are calculated;
wherein the second parameter set comprises at least one of: historical data of the virtual server required by the computing task corresponding to the total core number of the Central Processing Unit (CPU) and historical data of the load rate of the physical server;
the load layer module is configured to obtain a third parameter group, and determine third prediction data according to the second parameter group, the second prediction data, and the third parameter group, where the third prediction data includes at least one of the following: calculating the corresponding prediction data of the electric load required by the power task, the refrigeration facility and the charging facility;
wherein the third parameter set comprises at least one of: historical data of power loads, running state parameters, equipment aging rates, ambient temperatures, charging facility power supply loads and charging and discharging attenuation rates input by a server refrigeration facility, a machine room refrigeration facility and other production and living refrigeration facilities;
the integration module is used for integrating the first prediction data, the second prediction data and the third prediction data and determining integration information, and the integration information is used for determining optimized scheduling of the source network load storage unit.
2. The system of claim 1, wherein the new energy layer module comprises a wind-solar combined plant long-scale high-precision power prediction analysis unit and a centralized power monitoring analysis and high-precision prediction unit of a regional multi-wind-field plant cluster;
the wind and light combined station long-scale high-precision power prediction analysis unit is used for fitting the first parameter group to obtain first fitting information, and predicting the power generation power of a single wind and light combined station according to the first fitting information;
and the centralized power monitoring analysis and high-precision prediction unit of the regional multi-wind-field station cluster is used for fitting the first parameter group to obtain second fitting information, and predicting the power generation power of the wind-light combined cluster according to the second fitting information.
3. The system of claim 2, wherein the first fitting information corresponding to the wind-solar combined plant long-scale high-precision power prediction analysis subsystem is expressed by the following formula:
wherein, y a Historical and predictive data, x, representing the total generated electrical power of a wind farm a Historical and predicted data, x 'representing different meteorological parameters' a Historical and predicted data, x ″, representing different operation and maintenance parameters a Historical and predictive data representing the rate of decay of power generation by the plant.
4. The system of claim 2, wherein the second fitting information corresponding to the centralized power monitoring analysis and high-precision prediction subsystem of the regional multi-wind farm cluster is shown as follows:
wherein, y b Historical and predictive data, x, representing the total generated electrical power of a wind-solar cluster b Historical and predicted data, x 'representing different meteorological parameters' b Historical and predicted data, x ″, representing different operation and maintenance parameters b Historical period and expected life data, x ″, representative of different corporate sites' b Historical and predictive data representing the rate of decay of power generation by the plant.
5. The system of claim 1, wherein the computing layer module comprises a data centric computing power task predictive analysis unit;
and the data center computing power task prediction analysis unit is used for fitting the second parameter group to obtain third fitting information, and determining the second prediction data according to the third fitting information.
6. The system of claim 1, wherein the load layer modules comprise an internet technology equipment power load analysis and prediction analysis unit, a data center refrigeration facility power load analysis and prediction analysis unit, and a data center charging facility power load prediction analysis unit;
the internet technology equipment power load comprehensive analysis and prediction analysis unit is used for fitting the second prediction data and the second parameter group to obtain fourth fitting information, and determining the third prediction data according to the fourth fitting information;
the data center refrigeration facility power load comprehensive analysis and prediction analysis unit is used for fitting the second parameter group, the third parameter group and the second prediction data to obtain fifth fitting information, and determining prediction data of a machine room refrigeration facility and other production and living refrigeration facilities according to the fifth fitting information;
and the data center charging facility power load prediction analysis unit is used for fitting the third parameter group to obtain sixth fitting information, and determining prediction data of charging facilities in the park according to the sixth fitting information.
7. The system of claim 6, wherein the fourth fitting information corresponding to the internet technology equipment power load integrated analysis and predictive analysis subsystem is as follows:
y f ＝f(x f )
y g ＝f(x g )
wherein y is f 、y g Historical and predicted data representing the power loads required by a single power calculation task and all power calculation tasks in a short term and a medium-long term respectively, x f 、x g And the historical and predicted data respectively represent the total number of cores of the central processing unit CPU of the virtual server required by the short-term and medium-term single computing task and all the computing tasks.
8. The system of claim 6, wherein the fifth fitting information corresponding to the data center refrigeration facility power load integrated analysis and predictive analysis subsystem is as follows:
wherein y is h 、y i Respectively representing operating state parameters of the machine room refrigeration facility, historical and predicted data of the electrical load, y j 、y k Respectively representing the operating state parameters of other production and living refrigeration facilities, the history and prediction data of the power load, t h 、t j Respectively representing the historical and predicted time periods, x, corresponding to the operating condition parameters i 、x k Respectively representing the historical and predicted aging rates, x 'of equipment in the machine room refrigeration facility and other production and living refrigeration facilities' i 、x′ k Respectively representing history and prediction data x' of the environmental temperature corresponding to the machine room refrigeration facility and other production and living refrigeration facilities i Historical and predicted data representing physical server load rates.
9. The system of claim 6, wherein the sixth fitting information for the data center charging facility power load prediction analysis subsystem is as follows:
y l ＝f(t l )+f(x l )
wherein, y l Historical and predicted data representing charging facility power supply load, considering user average maximum power supply load, average power supply load for a particular date, t l Representing charging facility operating time period, x l History and prediction data representing charge and discharge decay rates.
10. A unit state active sensing and predicting method based on source network load storage is applied to a unit state active sensing and predicting analysis system and is characterized by comprising the following steps:
acquiring a first parameter group, and determining first prediction data according to the first parameter group, wherein the first prediction data comprises at least one of the following items: the power generation power of the wind-light combined station and the power generation power of the wind-light combined cluster;
wherein the first parameter set comprises at least one of: meteorological data, power generation attenuation rate and operation and maintenance parameters;
acquiring a second parameter group, and determining second prediction data according to the second parameter group, wherein the second prediction data comprises at least one of the following items: the total core number of a virtual server corresponding to a Central Processing Unit (CPU) and the load rate of a physical server required by the computing task are calculated;
wherein the second parameter set comprises at least one of: historical data of the virtual server required by the computing task corresponding to the total core number of the Central Processing Unit (CPU) and historical data of the load rate of the physical server;
acquiring a third parameter group, and determining third prediction data according to the second parameter group, the second prediction data and the third parameter group, wherein the third prediction data comprises at least one of the following data: calculating the corresponding prediction data of the electric load required by the power task, the refrigeration facility and the charging facility;
wherein the third parameter set comprises at least one of: historical data of power loads, running state parameters, equipment aging rates, ambient temperatures, charging facility power supply loads and charging and discharging attenuation rates input by a server refrigeration facility, a machine room refrigeration facility and other production and living refrigeration facilities;
and integrating the first prediction data, the second prediction data and the third prediction data and determining integration information, wherein the integration information is used for determining optimized scheduling of the source network load-storage units.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211197518.XA CN115481918A (en) | 2022-09-29 | 2022-09-29 | Active sensing and predictive analysis system for unit state based on source network load storage |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211197518.XA CN115481918A (en) | 2022-09-29 | 2022-09-29 | Active sensing and predictive analysis system for unit state based on source network load storage |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115481918A true CN115481918A (en) | 2022-12-16 |
Family
ID=84393836
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211197518.XA Pending CN115481918A (en) | 2022-09-29 | 2022-09-29 | Active sensing and predictive analysis system for unit state based on source network load storage |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115481918A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116683542A (en) * | 2023-07-26 | 2023-09-01 | 国网浙江省电力有限公司宁波供电公司 | Source network charge storage control method and device, computer equipment and storage medium |
| CN116865261A (en) * | 2023-07-19 | 2023-10-10 | 王克佳 | Power load prediction method and system based on twin network |
-
2022
- 2022-09-29 CN CN202211197518.XA patent/CN115481918A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116865261A (en) * | 2023-07-19 | 2023-10-10 | 王克佳 | Power load prediction method and system based on twin network |
| CN116865261B (en) * | 2023-07-19 | 2024-03-15 | 梅州市嘉安电力设计有限公司 | Power load prediction method and system based on twin network |
| CN116683542A (en) * | 2023-07-26 | 2023-09-01 | 国网浙江省电力有限公司宁波供电公司 | Source network charge storage control method and device, computer equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103390116B (en) | Use the photovoltaic power station power generation power forecasting method of stepping mode | |
| CN115481918A (en) | Active sensing and predictive analysis system for unit state based on source network load storage | |
| CN106228278A (en) | Photovoltaic power prognoses system | |
| CN114021420A (en) | Distributed photovoltaic ultra-short-term power generation prediction method and system | |
| CN113496311A (en) | Photovoltaic power station generated power prediction method and system | |
| CN109978242A (en) | The photovoltaic power generation cluster power forecasting method and device of scale are risen based on statistics | |
| CN115796393B (en) | Energy management optimization method, system and storage medium based on multi-energy interaction | |
| CN116865236A (en) | Medium-and-long-term power generation capacity prediction method and system based on new energy power generation | |
| CN112884601B (en) | Power system operation risk assessment method based on weather division strategy | |
| CN110909958A (en) | Short-term load prediction method considering photovoltaic grid-connected power | |
| CN115965132A (en) | Distributed photovoltaic digital twin system power prediction method based on GA-BP neural network | |
| CN115912351A (en) | Photovoltaic day-ahead model prediction method, system, equipment and medium | |
| CN111008727A (en) | Power distribution station load prediction method and device | |
| CN114021848A (en) | Generating capacity demand prediction method based on LSTM deep learning | |
| CN108346009A (en) | A kind of power generation configuration method and device based on user model self study | |
| CN109858668B (en) | Coordination prediction method for power load region in thunder and lightning climate | |
| CN117335411B (en) | Medium-and-long-term power generation capacity prediction method for photovoltaic power station group | |
| CN117669960A (en) | New energy power prediction method based on multivariable meteorological factors | |
| CN116484998A (en) | Distributed photovoltaic power station power prediction method and system based on meteorological similar day | |
| Hsu et al. | Short-term load forecasting by machine learning | |
| Zhou et al. | PV power characteristic modeling based on multi-scale clustering and its application in generation prediction | |
| CN114676866A (en) | Method, device and storage medium for correcting load prediction before day based on error correction | |
| CN115438839B (en) | Photovoltaic power ultra-short-term prediction system and method based on intelligent ammeter data | |
| CN117833300B (en) | Energy storage capacity configuration method based on intelligent power distribution equipment | |
| CN117595231B (en) | Intelligent power grid distribution management system and method thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |